In the second part of Implema Talks focusing on SAP BDC and data platforms, Susanne Söderholm and Johan Söderström explain two ways to connect SAP data to modern data platforms.
Either you bring non-SAP data into the SAP environment via SAP Business Data Cloud using modern data sharing instead of heavy batch jobs, or you share SAP data to an external data platform like Databricks or Snowflake, where analytics and AI are kept separate from the ERP systems.
Susanne points out that the external model often provides more flexibility with many data sources, multiple countries, or M&A, and that security, governance, and compliance are crucial for creating a single version of the truth and becoming AI-ready.
In this episode of Implema Talks, Johan Söderström and Susanne Söderholm take the next step in the series on data and analytics. The focus is on something many organizations are struggling with right now: how to connect SAP data with a modern data platform while simplifying architecture, cost, governance, and AI readiness. How does SAP BDC make things easier for SAP-focused companies?
Susanne Söderholm, Business Area Manager for BI, Data & Analytics at Implema, describes two main tracks that often work best in practice.
In short:
Two paths: Either you collect non-SAP data in the SAP environment via SAP Business Data Cloud, or you share SAP data to an external data platform.
An external data platform is often best when you have many sources, multiple countries, or M&A activity and want to keep analytics and AI separate from your ERP systems.
Modern data sharing replaces heavy batch jobs and makes it easier to manage access, security, and compliance while creating a single version of the truth.
Most companies no longer have a single “truth” in one system. SAP might be the core ERP, but data also exists in CRM, e-commerce, production, HR, logistics, IoT, partners, and external sources. As the demand for faster analysis, AI, and better decision support grows, the question becomes urgent: Where should the analytics capacity reside? In SAP or in an external data platform?
Susanne describes two main ways to build a modern data architecture.
This path often suits organizations that want to consolidate more within the SAP ecosystem and work from there. The point is to avoid heavy batch jobs and instead use modern sharing and smarter connections between sources and the platform.
This allows SAP to become a place where more data sources meet without building large, expensive, and hard-to-maintain integration chains.
The second track is to let SAP be one of several sources. You then perform analytics and AI in an external platform like Databricks or Snowflake, and eventually Microsoft Fabric when it fits the customer’s environment.
Susanne’s preference is clear here. In many cases, it’s simpler and more future-proof to separate analytics from the ERP systems as much as possible. This is especially true for organizations that acquire and sell companies or have multiple ERP systems over time. In those cases, you want to be able to “turn on or turn off” data without having to rebuild core systems every time the structure changes.
The conversation touches on a familiar problem. An ERP risks becoming overloaded with data points and logic—too much customization, too much “everything in one box.” This creates complexity and cost, and slows down the pace of change.
Johan describes it as being able to “dip into the data lake” to find the needle in the haystack. Transformation and refinement can be done once you know what you need, rather than as a massive undertaking before you even know which questions are most important.
Many companies already have tons of data stored in a data lake or data platform. When someone asks a business-critical question, you don’t need to first build a large, heavy, and perfect data model that “refines everything.” Instead, you can start by extracting a relevant slice of the data—”dipping into the data lake”—to quickly find signals, anomalies, or answers. That’s the “needle in the haystack.”
Once you know which questions are actually important and what data is needed, then you perform the heavy transformation and build more robust pipelines, models, and KPI definitions. The point is to avoid extensive preparation that costs time and money but risks optimizing the wrong things because you don’t yet know exactly what the business needs.
A key takeaway from the episode is the shift from classic serial copying to modern data sharing.
Instead of moving and duplicating data in multiple stages, you can often “view” data via controlled sharing. This can provide:
lower costs for data movement and storage
simpler architecture
faster time to insight
better conditions for AI readiness
Susanne also highlights an important aspect often missed in the discussion: security rules. Who gets to see what—at the row level, role level, and need-to-know level. This is crucial for maintaining compliance, ownership, and trust when sharing data across platforms.
Here is a practical rule of thumb based on the discussion in the video.
Choose Track 1, SAP Business Data Cloud as the hub, when:
you want to consolidate more within the SAP ecosystem
you have a clear SAP-centric governance and roadmap
you want to avoid heavy batch integration and modernize connections
Choose Track 2, external data platform, when:
you have many sources and want to be platform-neutral
you want to separate analytics and AI from the ERP systems
you have changing corporate structures, M&A, multiple ERPs, or need rapid adaptation
you want to build a “single version of the truth” that outlasts individual system changes
Want to know which path is right for you? SAP Business Data Cloud as a hub, or an external data platform for analytics and AI? Get in touch and we’ll run a short workshop where we map out data sources, use cases, governance, and your target vision.
What is the difference between bringing data into SAP and sharing data out of SAP?
Bringing data into SAP means the SAP environment becomes the hub. Sharing data out means SAP is one source and analytics take place in an external data platform.
Why do many want to separate analytics and AI from the ERP?
To reduce complexity in the ERP systems and make analytics more scalable, flexible, and easier to change when adding new companies or sources.
What needs to be secured when sharing data between platforms?
Governance, access rules, row-level security, compliance, and clear ownership. The right person should see the right data.
00:00 Intro. Data sources, data lakes, and simplifying the concepts
00:29 Susanne. Why data and analytics are always changing
00:55 SAP enters an already mature market
01:23 Databricks, Snowflake, and Microsoft: How to think about them
01:38 Two paths: Into the SAP environment or out to an external platform
02:40 Why analytics should often reside outside the ERP systems
03:32 The M&A scenario: Easier to turn data on and off
03:44 When the ERP becomes overloaded: Too many data points
04:31 SAP opens up: Cost and efficiency improve
04:56 AI readiness: Why Databricks and Snowflake are often a good fit
05:30 Security rules: Governance and proper access
05:52 Compliance and ownership: Why many hesitate
06:12 Wrap-up. Next episode on use cases
Get in touch with our team – we’re happy to help!
Whether you have a question, want to learn more about our solutions, or are simply curious how we can support your business – we’d love to hear from you.
